Retrieval method and system for property integrated management
Through the data middle platform and NL2SQL technology, the problems of information dispersion and data islands in the property management system are solved, natural language query is realized, query efficiency and accuracy are improved, user usage threshold is lowered, and cross-system retrieval services are provided.
Patent Information
- Application Number
- CN202411049844.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-01
- Publication Date
- 2024-12-03
AI Technical Summary
The existing property management system has dispersed information, incommunication of data, and inconsistent formats, which leads to users requiring professional knowledge to conduct cross-platform system query, which increases the threshold for users.
Using data middle platform and NL2SQL technology, we open up data islands in different systems, convert the natural language input by users into executable SQL statements, and realize cross-system data summary and query.
It lowers the threshold for users to use, improves query efficiency, diversity and accuracy, provides comprehensive cross-system user retrieval services, greatly simplifying the query process of property data.
Smart Images

Figure CN119066087A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of integrated property management, and in particular to a retrieval method and system for integrated property management. Background Art
[0002] With the development of science and technology and the increasing complexity of property management business, management needs have shown obvious new characteristics. Property information is scattered in multiple information channels, such as real estate management systems, rental management systems, contract management systems and other platform systems. Information is not interoperable and has inconsistent formats, resulting in information fragmentation.
[0003] At this time, if users want to obtain comprehensive and accurate property management information, they need to have corresponding professional knowledge, otherwise they need to visit multiple systems one by one to query. Sometimes querying one system cannot meet the needs of users, and they need to query results from different systems. Therefore, traditional property management retrieval methods require users to have certain professional knowledge. User queries need to use specific keywords or structured query tags, and cross-platform system queries require professional knowledge, resulting in a high threshold for users. Summary of the invention
[0004] The purpose of the present invention is to overcome the shortcomings of the prior art. The present invention provides a retrieval method and system for comprehensive property management. The data of each property information management system is collected based on the data middle platform, and the data islands between different systems and different databases are connected. Then, the NL2SQL technology is combined with the data middle platform to allow the natural language input by the user to be converted into executable query statements, so that the data in the data middle platform can be easily and efficiently queried. With the help of the data middle platform and NL2SQL technology, this method can not only connect the data of different systems and different databases of the property, effectively summarize and collect the property data, and provide comprehensive cross-system user retrieval services, but also improve the data quality of multiple systems, improve the user's query efficiency, diversity and accuracy, and only need to use natural language description to input the content you want to query, which greatly reduces the user's usage threshold.
[0005] In order to solve the above technical problems, an embodiment of the present invention provides a retrieval method for comprehensive property management, which is applied to the retrieval scenario of comprehensive property management;
[0006] The retrieval method for comprehensive property management comprises:
[0007] Collect data from various property information management systems based on the data center;
[0008] In the data center, a search library is constructed based on the data of each property information management system;
[0009] Collect the search content entered by the user and convert it into SQL statements through NL2SQL semantic analysis;
[0010] Determine the search results based on the SQL statement and the search library.
[0011] Optionally, the data of each property information management system is collected based on the data center, including:
[0012] Collecting data from various property information management systems, the property information management system including at least one of the asset management systems in the housing management system, the rental management system, and the contract management system;
[0013] Linking data from various property information management systems;
[0014] In the data center, data from various property information management systems are collected and processed;
[0015] The data of different property information management systems are divided into ODS original data, DWD detailed data, DWS aggregated data and ADS business data.
[0016] Optionally, the data collection of each property information management system based on the data center platform also includes:
[0017] Perform data preprocessing and data cleaning based on data from various property information management systems;
[0018] The format of the data output after data preprocessing is converted and the data formats of each data are unified.
[0019] Optionally, in the data center, a search library is constructed according to the data of each property information management system, including:
[0020] In the data center, locate the data of each property information management system;
[0021] Perform cross-system or cross-table association on the data of each property information management system to associate the data of each property information management system;
[0022] A search library is constructed based on the association status of data in each property information management system.
[0023] Optionally, collecting the search content input by the user and converting it into an SQL statement through NL2SQL semantic analysis includes:
[0024] Collecting search content input by the user, where the search content input by the user includes text content or voice content;
[0025] Freeze the search content entered by the user;
[0026] In NL2SQL semantic analysis, the search content entered by the user is converted into SQL statements.
[0027] Optionally, the collecting of search content input by the user and converting it into an SQL statement through NL2SQL semantic analysis also includes:
[0028] Get the search content entered by the user;
[0029] Extract key information of the searched content as feature information expressing the meaning of the content;
[0030] According to the key feature information, the minimum word segmentation unit is obtained;
[0031] In the overall sentence analysis, key feature information is divided into paragraphs, key words are extracted and connected, and the sentence structure is adjusted according to the search content entered by the user.
[0032] Optionally, the collecting of search content input by the user and converting it into an SQL statement through NL2SQL semantic analysis also includes:
[0033] Combine semantics and analyze the interrelationships of contexts to optimize and correct the search content entered by the user to obtain the natural language text that is ultimately required for the query;
[0034] The natural language is used as the search condition and converted into an executable SQL query statement through NL2SQL technology.
[0035] Optionally, the collecting of search content input by the user and converting it into an SQL statement through NL2SQL semantic analysis also includes:
[0036] Assume that the natural language of the user's query is Q, and the generated SQL query statement is S;
[0037] Probability calculation formula: P(S|Q)=P(Q|S)·P(S)÷P(Q)
[0038] in:
[0039] P(S|Q): the probability that the generated SQL query statement is S given the user query Q;
[0040] P(Q|S): the probability that the user's query is Q given the generated SQL query S;
[0041] P(S): the prior probability that the generated SQL query statement is S;
[0042] P(Q): the prior probability that the user query is Q;
[0043] This formula is calculated using the following steps:
[0044] Calculate P(Q|S): Use the language model to calculate the probability that the user query is Q given the generated SQL query S.
[0045] Calculate P(S): Based on the training data or domain knowledge, calculate the prior probability that the generated SQL query statement is S;
[0046] Calculate P(Q): Calculate the prior probability that the user query is Q based on the training data or the user's historical queries;
[0047] Calculate P(S|Q): Using the above calculation results, get the probability that the final generated SQL query statement is S.
[0048] Optionally, determining the search result according to the SQL statement and the search library includes:
[0049] Collect SQL statements;
[0050] Associate SQL statements and search libraries;
[0051] Determine the search results based on the SQL statement and the search library.
[0052] In addition, an embodiment of the present invention further provides a retrieval system for comprehensive property management, the retrieval system for comprehensive property management comprising:
[0053] The collection module is used to collect data from various property information management systems based on the data center;
[0054] The search library module is used to build a search library in the data center according to the data of each property information management system;
[0055] The statement module is used to collect the search content entered by the user and convert it into SQL statements through NL2SQL semantic analysis;
[0056] The search result module is used to determine the search results based on the SQL statement and the search library.
[0057] In an embodiment of the present invention, through the method in the embodiment of the present invention, data of various property information management systems are collected based on the data middle platform; in the data middle platform, a retrieval library is constructed according to the data of various property information management systems; the retrieval content input by the user is collected and converted into SQL statements through NL2SQL semantic analysis; the retrieval results are determined according to the SQL statements and the retrieval library. At this time, data of various property information management systems are collected based on the data middle platform, and data islands between different systems and different databases are connected. Then, NL2SQL technology is used in combination with the data middle platform to allow the natural language input by the user to be converted into executable query statements, so that the data in the data middle platform can be easily and efficiently queried. With the help of the data middle platform and NL2SQL technology, this method can not only connect the data of different databases of different property systems, effectively summarize and collect property data, and provide comprehensive cross-system user retrieval services, but also improve the data quality of multiple systems, improve the query efficiency, diversity and accuracy of users, and only need to use natural language description to input the content you want to query, which greatly reduces the user's usage threshold. BRIEF DESCRIPTION OF THE DRAWINGS
[0058] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0059] Figure 1 is a flow chart of a retrieval method for comprehensive property management in an embodiment of the present invention;
[0060] Figure 2 is a flow chart of S11 in the retrieval method for comprehensive property management in an embodiment of the present invention;
[0061] Figure 3 is a flow chart of S12 in the retrieval method for comprehensive property management in an embodiment of the present invention;
[0062] Figure 4 is a flow chart of S13 in the retrieval method for comprehensive property management in an embodiment of the present invention;
[0063] Figure 5 is a flow chart of S14 in the retrieval method for comprehensive property management in an embodiment of the present invention;
[0064] Figure 6 It is a schematic diagram of the structure of the retrieval system for comprehensive property management in an embodiment of the present invention;
[0065] Figure 7The figure is a hardware diagram of an electronic device according to an exemplary embodiment. DETAILED DESCRIPTION
[0066] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0067] Example
[0068] See also Figures 1 to 7 , a retrieval method for comprehensive property management, applied to the retrieval scenario of comprehensive property management; the retrieval method for comprehensive property management includes:
[0069] Step S11: Collecting data from various property information management systems based on the data center;
[0070] Step S12: In the data center, a search library is constructed based on the data of each property information management system;
[0071] Step S13: collecting the search content input by the user and converting it into SQL statements through NL2SQL semantic analysis;
[0072] Step S14: Determine the search result according to the SQL statement and the search library.
[0073] In an embodiment of the present invention, through the method in the embodiment of the present invention, data from various property information management systems are collected based on the data middle platform, and data islands between different systems and different databases are connected. Then, NL2SQL technology is used in combination with the data middle platform to allow the natural language input by the user to be converted into executable query statements, so that the data in the data middle platform can be easily and efficiently queried. With the help of the data middle platform and NL2SQL technology, this method can not only connect the data of different databases of different property systems, effectively summarize and collect property data, and provide comprehensive cross-system user retrieval services, but also improve the data quality of multiple systems, improve the user's query efficiency, diversity and accuracy, and only need to use natural language description to enter the content you want to query, which greatly reduces the user's usage threshold.
[0074] In step S11, data from various property information management systems are collected based on the data center;
[0075] In the specific implementation process of the present invention, the specific steps may be:
[0076] S111: Collecting data from various property information management systems, where the property information management system includes at least one of an asset management system in a housing management system, a rental management system, and a contract management system;
[0077] S112: Associating data from various property information management systems;
[0078] S113: In the data center, data from various property information management systems are aggregated and processed;
[0079] S114: dividing the data of different property information management systems into ODS original data, DWD detailed data, DWS aggregated data and ADS business data;
[0080] S115: performing data preprocessing and data cleaning based on data from various property information management systems;
[0081] S116: Convert the format of the data output after data preprocessing, and unify the data formats of each data.
[0082] In an embodiment of the present application, data from each property information management system is collected, and the property information management system includes at least one of an asset management system including a house management system, a rental management system, and a contract management system. At this time, the data from each property information management system is managed and controlled, thereby linking the data from each property information management system.
[0083] Furthermore, a data middle platform was introduced, in which the data of each property information management system were collected and processed, so that the data of different property information management systems were divided into ODS original data, DWD detailed data, DWS aggregated data and ADS business data. At this time, the data islands between different systems and databases were connected, and then data governance technology was used to clean, integrate, convert, etc. a large amount of potentially duplicate and chaotic data, so as to improve data quality and unify data standards.
[0084] Therefore, data preprocessing and data cleaning are performed based on the data of each property information management system; the format of the data output after data preprocessing is converted, and the data format of each data is unified.
[0085] Specifically, the data preprocessing and cleaning module is used to preprocess and clean the comprehensive property management data of the scheduling process, such as deleting records with missing important fields, filling "missing content" with non-numeric optional missing values, etc.
[0086] Data conversion is used to convert the comprehensive property management data of the scheduling process into a unified data format standard. For example, the amount data format is unified as Numer i cal, retaining two decimal places; the date data format is unified as "YYYY-MM-DD HH:MM:SS" format.
[0087] At the same time, data scheduling ETL (Extract-Transform-Load) technology is used to establish scheduling tasks for data in multiple property information management systems through direct connection with HTTP / HTTPS interfaces and direct connection with databases, and the scheduling frequency is set. For data with high real-time requirements, scheduling is set to be by seconds, and for data with low real-time requirements, scheduling is set to be by days. In addition, scheduling log monitoring is added, and multiple scheduling can be performed for data with scheduling errors to ensure correct data synchronization.
[0088] In step S12, in the data center, a search library is constructed based on the data of each property information management system;
[0089] In the specific implementation process of the present invention, the specific steps may be:
[0090] S121: Locate each property information management system data in the data center;
[0091] S122: Perform cross-system or cross-table association on the data of each property information management system to associate the data of each property information management system;
[0092] S123: Construct a search library based on the association status of data in each property information management system.
[0093] In an embodiment of the present application, each property information management system data is located in the data center to facilitate cross-system or cross-table association of each property information management system data to associate each property information management system data, thereby building a retrieval library based on the association status of each property information management system data.
[0094] At the same time, the data association module is used to associate the comprehensive property management data of the scheduling process across systems or tables to reduce data redundancy. For example, the house management table, rental management table and contract management table are associated across systems to improve the efficiency of subsequent queries.
[0095] Furthermore, the data integration and deduplication module is used to deduplicate the property management data of the scheduling process, remove duplicate records in the data, and reduce duplicate records of the same information in different systems. For example, the full amount of house management information is retained, and the fields in the house rental system that are repeated with the house management system are removed, and only the associated house ID field is retained, etc. Optionally, the database storage technology is used to divide the different system data into ODS original data, DWD detailed data, DWS aggregated data and ADS business data, and store them in the relational database PostgreSQL.
[0096] Optionally, for the data governance process, from data preprocessing and cleaning, data conversion, data association and data integration to deduplication, set missing value fields through different SQL statements, unify data format standards, and use processing statements to associate across systems and tables, as well as deduplication of data, reduce data redundancy, and improve data query efficiency and data accuracy. And use database storage technology to divide different system data into ODS original data, DWD detailed data, DWS aggregated data and ADS business data, and store them in the relational database PostgreSQL.
[0097] In step S13, the search content input by the user is collected and converted into SQL statements through NL2SQL semantic analysis;
[0098] In the specific implementation process of the present invention, the specific steps may be:
[0099] S131: collecting search content input by the user, where the search content input by the user includes text content or voice content;
[0100] S132: freeze the search content input by the user;
[0101] S133: In NL2SQL semantic analysis, the search content input by the user is converted into an SQL statement.
[0102] In an embodiment of the present application, the search content input by the user is collected, and the search content input by the user includes text content or voice content, so as to control according to the search content input by the user, thereby freezing the search content input by the user and converting the search content input by the user into an SQL statement.
[0103] Therefore, the search content input by the user is obtained; the key information of the search content is extracted as the feature information expressing the meaning of the content; the minimum word segmentation unit is obtained according to the key feature information; in the overall sentence analysis, the key feature information is divided into paragraphs, the keywords are taken out and connected, and the composition of the sentence is adjusted according to the search content input by the user; the search content input by the user is optimized and corrected by combining the semantics and analyzing the mutual correlation of the context to obtain the natural language text that needs to be queried in the end; the natural language is used as the search condition and converted into an executable SQL query statement through the NL2SQL technology.
[0104] The present invention provides a property comprehensive management retrieval method based on data middle platform and NL2SQL. Compared with the current retrieval technology using key tags and keywords, this method significantly improves the accuracy and professionalism of property comprehensive management data retrieval. Users do not need to have professional knowledge, and the threshold for use is greatly reduced, thereby further narrowing the distance between users and data from different property information management systems. Users can more freely retrieve more information across systems and data tables, and realize retrieval intentions through richer natural language input methods according to personal needs. This not only greatly reduces the cumbersomeness of current technical solutions, but also liberates programmers from changing management needs. This method can well make up for the shortcomings of the existing technology and improve the application capabilities of structured and unstructured texts in real natural language retrieval landing scenarios. Furthermore, the value of property management data is more fully and deeply excavated, providing non-professional users with help to obtain effective retrieval information at the fastest speed, thereby enhancing the user's sense of gain.
[0105] In addition, the present invention adopts the mainstream AST syntax tree to establish the table-level blood relationship in data governance. Perform lexical analysis and syntax analysis on Hive SQL in the AST syntax tree to obtain the corresponding AST original abstract syntax tree; then use the AST syntax tree pruning optimization to reduce the number of traversals and improve the efficiency of semantic parsing. Specifically, two aspects of optimization are made: delete the invalid parsing nodes involved in the token, such as order by, differed by, class by, sort by and limit; for the subquery of where / have ng in the token, under the premise of ensuring the correctness of SQL syntax and semantic integrity, adopt the 1=1 equivalent strategy for equivalent replacement, reducing the complexity of blood relationship parsing; through the above two pruning operations, the complexity of SQL statements can be reduced, and the level of AST syntax tree can be reduced, further reducing the number of recursive traversals of AST tree, reducing the complexity of blood relationship analysis, and improving statement parsing efficiency. Then traverse the AST to obtain the upstream table name and downstream table name, and most of the SQL statement fragments, namely CTE, exist in the SQL statement. Since it does not play a key role in blood relationship analysis and brings great trouble to SQL analysis, blood relationship analysis needs to identify the CTE type and replace and delete it. This processing method allows NL2SQL technology to search more accurately and quickly when using natural language search queries.
[0106] In addition, the collecting of the search content input by the user and converting it into SQL statements through NL2SQL semantic analysis also includes:
[0107] Assume that the natural language of the user's query is Q, and the generated SQL query statement is S;
[0108] Probability calculation formula: P(S|Q)=P(Q|S)·P(S)÷P(Q)
[0109] in:
[0110] P(S|Q): the probability that the generated SQL query statement is S given the user query Q;
[0111] P(Q|S): the probability that the user's query is Q given the generated SQL query S;
[0112] P(S): the prior probability that the generated SQL query statement is S;
[0113] P(Q): the prior probability that the user query is Q;
[0114] This formula is calculated using the following steps:
[0115] Calculate P(Q|S): Use the language model to calculate the probability that the user query is Q given the generated SQL query S.
[0116] Calculate P(S): Based on the training data or domain knowledge, calculate the prior probability that the generated SQL query statement is S;
[0117] Calculate P(Q): Calculate the prior probability that the user query is Q based on the training data or the user's historical queries;
[0118] Calculate P(S|Q): Using the above calculation results, get the probability that the final generated SQL query statement is S.
[0119] Such a probability calculation formula can be used in NL2SQL to better evaluate the quality of generated SQL query statements and help select the query statements that are most likely to match the user's query intent.
[0120] S14: Determine the search result according to the SQL statement and the search library;
[0121] In the specific implementation process of the present invention, the specific steps may be:
[0122] S141: Collect SQL statements;
[0123] S142: Associating SQL statements and search libraries;
[0124] S143: Determine the search result based on the SQL statement and the search library.
[0125] In an embodiment of the present application, SQL statements are collected so as to associate the SQL statements with a search library, thereby determining search results based on the SQL statements and the search library.
[0126] At this point, users can more freely retrieve more information across systems and data tables, and realize search intentions through richer natural language input methods according to personal needs. This not only greatly reduces the cumbersomeness of current technical solutions, but also liberates programmers from facing changing management needs. This method can well make up for the shortcomings of existing technologies and improve the application capabilities of structured and unstructured texts in real natural language retrieval scenarios. Furthermore, the value of property management data is more fully and deeply explored, providing non-professional users with help in obtaining effective search information at the fastest speed, thereby enhancing the user's sense of gain.
[0127] At this time, the data of each property information management system is collected based on the data middle platform; in the data middle platform, a search library is built according to the data of each property information management system; the search content entered by the user is collected and converted into SQL statements through NL2SQL semantic analysis; the search results are determined according to the SQL statements and the search library. At this time, the data of each property information management system is collected based on the data middle platform, and the data islands between different systems and databases are connected. Then, NL2SQL technology is used in combination with the data middle platform to allow the natural language entered by the user to be converted into executable query statements, so that the data in the data middle platform can be easily and efficiently queried. With the help of the data middle platform and NL2SQL technology, this method can not only connect the data of different systems and databases of the property, effectively summarize and collect the property data, and provide comprehensive cross-system user search services, but also improve the data quality of multiple systems, improve the user's query efficiency, diversity and accuracy, and only need to use natural language description to enter the content you want to query, which greatly reduces the user's usage threshold.
[0128] Example
[0129] See also Figure 6 , Figure 6 It is a schematic diagram of the structural composition of the retrieval system for comprehensive property management in an embodiment of the present invention.
[0130] like Figure 6 As shown, a retrieval system for comprehensive property management, the retrieval system for comprehensive property management includes:
[0131] The collection module 21 is used to collect data from various property information management systems based on the data center;
[0132] The search library module 22 is used to construct a search library in the data center according to the data of each property information management system;
[0133] The statement module 23 is used to collect the search content input by the user and convert it into SQL statements through NL2SQL semantic analysis;
[0134] The search result module 24 is used to determine the search result according to the SQL statement and the search library.
[0135] Example
[0136] See also Figure 7 , refer to the following Figure 7 The electronic device 40 according to this embodiment of the present invention will be described. Figure 7 The electronic device 40 shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.
[0137] like Figure 7 As shown, the electronic device 40 is in the form of a general computing device. The components of the electronic device 40 may include but are not limited to: at least one processing unit 41, at least one storage unit 42, and a bus 43 connecting different system components (including the storage unit 42 and the processing unit 41).
[0138] The storage unit stores program codes, which can be executed by the processing unit 41, so that the processing unit 41 executes the steps according to various exemplary embodiments of the present invention described in the above “Embodiment Method” section of this specification.
[0139] The storage unit 42 may include a readable medium in the form of a volatile storage unit, such as a random access memory unit (RAM) 421 and / or a cache memory unit 422 , and may further include a read-only memory unit (ROM) 423 .
[0140] The storage unit 42 may also include a program / utility 424 having a set (at least one) of program modules 425, such program modules 425 including but not limited to: an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include an implementation of a network environment.
[0141] Bus 43 may represent one or more of several types of bus structures, including a memory unit bus or memory unit controller, a peripheral bus, an accelerated graphics port, a processing unit, or a local bus using any of a variety of bus architectures.
[0142] The electronic device 40 may also communicate with one or more external devices (e.g., keyboards, pointing devices, Bluetooth devices, etc.), one or more devices that enable a user to interact with the electronic device 40, and / or any device that enables the electronic device 40 to communicate with one or more other computing devices (e.g., routers, modems, etc.). Such communication may be performed via an input / output (I / O) interface 44. Furthermore, the electronic device 40 may also communicate with one or more networks (e.g., a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) via a network adapter 45. Figure 7 As shown, the network adapter 45 communicates with other modules of the electronic device 40 via the bus 43. It should be understood that although Figure 7 Not shown, other hardware and / or software modules may be used in conjunction with the electronic device 40, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup planning systems.
[0143] Through the description of the above implementation, it is easy for those skilled in the art to understand that the example implementation described here can be implemented by software, or by software combined with necessary hardware. Therefore, the technical solution according to the implementation of the present disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, including several instructions to enable a computing device (which can be a personal computer, a server, a terminal device, or a network device, etc.) to execute the method according to the implementation of the present disclosure.
[0144] Those skilled in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium, and the storage medium can include: a read-only memory (ROM), a random access memory (RAM), a disk or an optical disk, etc. In addition, it stores computer program instructions, and when the computer program instructions are executed by a computer, the computer executes the above method.
[0145] In addition, the above is a detailed introduction to the retrieval method and system for comprehensive property management provided by the embodiments of the present invention. Specific examples are used herein to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea; at the same time, for those skilled in the art, according to the ideas of the present invention, there will be changes in the specific implementation methods and application scopes. In summary, the content of this specification should not be understood as a limitation on the present invention.
Claims
1. A retrieval method for comprehensive property management, characterized in that: Applied to retrieval scenarios for comprehensive property management; The retrieval method for comprehensive property management comprises: Collect data from various property information management systems based on the data center; In the data center, a search library is constructed based on the data of each property information management system; Collect the search content entered by the user and convert it into SQL statements through NL2SQL semantic analysis; Determine the search results based on the SQL statement and the search library.
2. The retrieval method for comprehensive property management according to claim 1, characterized in that: The data center-based collection of data from various property information management systems includes: Collecting data from various property information management systems, the property information management system including at least one of the asset management systems in the housing management system, the rental management system, and the contract management system; Linking data from various property information management systems; In the data center, data from various property information management systems are collected and processed; The data of different property information management systems are divided into ODS original data, DWD detailed data, DWS aggregated data and ADS business data.
3. The retrieval method for comprehensive property management according to claim 2, characterized in that: The data center-based collection of data from various property information management systems also includes: Perform data preprocessing and data cleaning based on data from various property information management systems; The format of the data output after data preprocessing is converted and the data formats of each data are unified.
4. The retrieval method for comprehensive property management according to claim 1, characterized in that: In the data center, a search library is constructed based on the data of each property information management system, including: In the data center, locate the data of each property information management system; Perform cross-system or cross-table association on the data of each property information management system to associate the data of each property information management system; A search library is constructed based on the association status of data in each property information management system.
5. The retrieval method for comprehensive property management according to claim 1, characterized in that: The collecting of the search content input by the user and converting it into SQL statements through NL2SQL semantic analysis includes: Collecting search content input by the user, where the search content input by the user includes text content or voice content; Freeze the search content entered by the user; In NL2SQL semantic analysis, the search content entered by the user is converted into SQL statements.
6. The retrieval method for comprehensive property management according to claim 5, characterized in that: The collecting of the search content input by the user and converting it into SQL statements through NL2SQL semantic analysis also includes: Get the search content entered by the user; Extract key information of the searched content as feature information expressing the meaning of the content; According to the key feature information, the minimum word segmentation unit is obtained; In the overall sentence analysis, key feature information is divided into paragraphs, key words are extracted and connected, and the sentence structure is adjusted according to the search content entered by the user.
7. The retrieval method for comprehensive property management according to claim 6, characterized in that: The collecting of the search content input by the user and converting it into SQL statements through NL2SQL semantic analysis also includes: Combine semantics and analyze the interrelationship of contexts to optimize and correct the search content entered by the user to obtain the natural language text that is ultimately required for the query; The natural language is used as the search condition and converted into an executable SQL query statement through NL2SQL technology.
8. The retrieval method for comprehensive property management according to claim 7, characterized in that: The collecting of the search content input by the user and converting it into SQL statements through NL2SQL semantic analysis also includes: Assume that the natural language of the user's query is Q, and the generated SQL query statement is S; Probability calculation formula: P(S|Q)=P(Q|S)·P(S)÷P(Q) in: P(S|Q): the probability that the generated SQL query statement is S given the user query Q; P(Q|S): the probability that the user's query is Q given the generated SQL query S; P(S): the prior probability that the generated SQL query statement is S; P(Q): the prior probability that the user query is Q; This formula is calculated using the following steps: Calculate P(Q|S): Use the language model to calculate the probability that the user query is Q given the generated SQL query S. Calculate P(S): Based on the training data or domain knowledge, calculate the prior probability that the generated SQL query statement is S; Calculate P(Q): Calculate the prior probability that the user query is Q based on the training data or the user's historical queries; Calculate P(S|Q): Using the above calculation results, get the probability that the final generated SQL query statement is S.
9. The retrieval method for comprehensive property management according to claim 8, characterized in that: Determining the search results according to the SQL statement and the search library includes: Collect SQL statements; Associate SQL statements and search libraries; Determine the search results based on the SQL statement and the search library.
10. A retrieval system for comprehensive property management, characterized in that: The retrieval system for comprehensive property management is applied to the retrieval method for comprehensive property management as claimed in any one of claims 1 to 9, and the retrieval system for comprehensive property management comprises: The collection module is used to collect data from various property information management systems based on the data center; The search library module is used to build a search library in the data center according to the data of each property information management system; The statement module is used to collect the search content entered by the user and convert it into SQL statements through NL2SQL semantic analysis; The search result module is used to determine the search results based on the SQL statement and the search library.
